Cross-lingual semantics for crime analysis using associate constraint network

Abstract

In light of the Bali bombings, East Asia nations rally the international community into a broad-based coalition to support a war against terrorism. The information sharing among different countries provides a challenge for cross-lingual semantic interoperability. In this work, we model the problem as an associate constraint network and propagation by backmarking is proposed for creating the cross-lingual concept space. The approach deals with structured as well as unstructured data, addresses relevancy of information, offers the user with associative navigation through the information embedded in the database, enables conduction of multiple languages. Evidence is presented to show that a constraint programming approach performs well and has the advantage in terms of tractability, ordering and efficiency over the Hopfield network. The research output consisted of a thesaurus-like, semantic network knowledge base relied on statistical correlation analysis of the semantics embedded in the documents of English/Chinese parallel corpus. © Springer-Verlag Berlin Heidelberg 2004.